基于多点映射分解的网络突变流量分解仿真  

Based on the Multi-Point Mapping Decomposition Flow Decomposition Simulation Network Mutations

作　　者：李亚[1,2] 刘伟[1] 

机构地区：[1]周口师范学院计算机科学与技术学院,河南周口466001 [2]北京交通大学信息系统研究所,北京100044

出　　处：《计算机仿真》2013年第9期298-301,共4页Computer Simulation

基　　金：国家自然科学基金项目(61103143);河南省重点科技攻关项目(102102210265);河南省基础与前沿研究项目(132300410400)

摘　　要：关于无线网络通信流量优化管理问题,当前网络在遭受到异常攻击,同时由于大规模登录时,流量会发生短时间巨幅变化,使得单个节点面临瘫痪风险。传统的网络调度算法是基于单个节点的分流完成的,一旦流量变化巨大,会造成分流失败,引起网络瘫痪。提出了一种多任务多点映射分解算法的网络突变流量分解管理方法。将Map函数与Reduce函数相结合,能够将网络特征构成的数据集合传递到分布式网络文件系统中,从而降低网络文件需要耗费的存储时间。将网络流量突然增大的相关特征与数据库中的任务进行分解处理,能够获取大量的子任务,完成网络突变流量的分解管理。实验结果表明,改进算法进行网络突变流量分解管理,能够极大地提高网络突变流量分解管理方法的性能,提高网络通信效率。Current network in abnormal to attack,or large-scale login at the same time,huge change flow will occur for short periods of time,make the paralysis risk faced by a single node.The current network scheduling algorithm is based on a single node shunt,once a great changing flow,can cause shunt failure,cause the network down.A multitasking mutation multi-point mapping decomposition algorithm of the network flow decomposition approach to management.Combining the Map function and a Reduce function,to be able to transfer the data collection of the network characteristics to the distributed network file system,so as to Reduce the storage time of the network file requires.Will be related to the characteristics of network traffic increases suddenly against a database of task decomposition processing,able to obtain large amounts of subtasks,final mutation flow decomposition of network management.Experimental results show that the algorithm presented in this paper for network mutation decomposition of traffic management,can greatly improve the performance of network mutation management flow decomposition method,so as to provide more high-quality network services.

关 键 词：分解管理 网络流量 匹配处理 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程]